Data is the raw material of insurance – unwise limits on its use would weaken insurance cover

Data is the single most valuable resource in insurance business. In the last two decades, the volume of data has grown beyond the wildest dreams of risk analysts. More data means more accurate risk assessment and higher-quality insurance products for customers.

With accurate and extensive data, insurers can improve risk modelling and provide customers with more diverse options for insurance cover. Accurately modelled risks are also easier to insure.

The logical apex of data use is artificial intelligence or AI, which enables access to an even greater volume of relevant data. But the risks of ever-growing volumes of data are numerous, especially if the data is AI-generated. AI regulation is therefore an absolute necessity. The EU AI Act applied in August 2024 encapsulates the key points well: it requires that all AI systems developed or used in the EU must be trustworthy and human-centric, fundamental rights must be protected and rules for the use of AI harmonised.

When it comes to the insurance sector’s valuable data, legislators need to strike a careful balance. Restricting data access and usage will limit product development and hamper insurance companies’ ability to insure and set prices especially for new risk types. For customers to benefit from these new products, data use restrictions must not be proposed without a comprehensive understanding of the use of data in the insurance sector.

On the other hand, the application of AI may lead to unfairness and discrimination if the system identifies correlation between variables that in reality have no causal relationship. Preventing this from happening requires transparency, quality data and human oversight, all of them areas in which insurance companies are continuously working to improve.

======
Regulation must be able to change with the times
if the future is different from what we now imagine it to be.
======

It’s worth reiterating: data has always been the insurance sector’s most valuable resource, and access to it must not be unwisely limited. Moreover, many of the risks related to AI and data are already taken into account in insurance sector regulation, which is technology-agnostic and also applies to the use of artificial intelligence. The advantages and benefits of AI are cancelled out if the same information must be reported multiple times under two separate regulatory frameworks.

It is important that regulation – especially when it concerns AI and the use of insurance data – is not based solely on forecasts but as strongly on factual information as possible.

Regulation that is principles-based and appropriately sector-specific stands the test of time better than overly detailed regulation. Legal practice provides a good means to adjust and adapt to changing outlooks and the eventuality that the future is different from what we imagined.